Researchers show Twitter is far faster than the police at predicting riots


After analyzing more than 1.6 million tweets, researchers at Cardiff University found that Twitter could identify dangerous situations far faster than standard police reports.

The researchers came to their findings after analyzing tweets sent during the 2011 London Riots, according to the Verge. The team found the tweets using a series of machine learning algorithms, which automatically scanned Twitter to identify potential threats. As Gizmodo explained, the researchers then took the tweets and added variables like location, time sent and hashtags. With the variables the researchers were able to conclude that tweets about disruptive events were sent minutes to nearly an hour before area police were alerted.

"In this research, we show that online social media are becoming the go-to place to report observations of everyday occurrences — including social disorder and terrestrial criminal activity," Pete Burnap, co-author of the study, told CNet.

And while being able to better track violent crime in real time sounds like a good thing, the technology comes with a potentially high cost.

"Social media was once incredibly effective at helping dissidents communicate under authoritarian regimes, as we saw during the Arab Spring," Cullen Hoback, director of the documentary Terms and Conditions May Apply, said via email. "However, governments eventually figured out how to monitor and undermine those systems. Now, law enforcement agencies routinely monitor social networks for key words and phrases to 'prevent' crime — this is nothing new."

Hoback added that police departments using social media to respond more quickly to emergencies seems like a logical evolution of the surveillance state.

"Yes, we want firefighters to be able to respond more quickly," he said. "No, we don't want police silencing protests before they happen."

And silencing protestors using predictive social media has already become a proven problem.

As Mic reported in 2016, social media giants Facebook, Twitter and Instagram all gave Geofeedia, a location-based social media monitoring app, access to user information. That information was then passed on by Geofeedia to police who had "special access" to the social media sites and used it to target activists of color, according to the ACLU.

"The potential downside of observing social media generally [is] there is an opportunity to profile individuals," Burnap told Gizmodo, "which we’re avoiding in this particular case by doing aggregate-level analysis and not focus on the individual."

So how can you protect your privacy and rights online? According to Hoback, people can opt for a paid virtual private network service, or VPN, for their cell phone and home internet. He added that since internet service providers are are now allowed to sell all of your online activity, using a VPN is truly the only way to protect yourself. A VPN, like Tunnelbear, costs about $50 a year, according to Hoback. Want a totally free trick? Simple. As Hoback said, "Don't post anything online you don't want the whole world to see."